The goal of this project is to discover the dynamical principles and mechanisms at play both within and between human brains during real-time social interaction. The research plan employs a three-pronged approach that combines (1) experimental manipulations to test specific hypotheses regarding key issues in the neurophysiology of social neuroscience (2) sophisticated measurement and analysis tools from the theory of dynamical systems, including virtual partner interaction (behavioral dynamic clamp of reciprocally coupled humans and model-partners) and (3) multiscale neurocomputational modeling of both structure and function in order to advance our understanding of how individual behavior and the interaction of individuals drives basic forms of social behavior. In our previous research, we established a comprehensive framework to tackle real- time interactions between people in simple, well-defined experimental paradigms in which pairs of participants simultaneously performed and perceived each other's movements. The research program led to the discovery of the phi complex, a neuromarker of social coordination. Also clarified were the contributions of other neuromarkers, especially alpha and mu, to different phases and facets of social behavior. What is most needed now -and what we seek support for in the present Competing Renewal- is to understand the dynamical orchestration of identified neuromarkers over the course of social behavior. The experimental thread works hand-in-hand with neurocomputational modeling of social behavior, theoretical models informing experiments and vice-versa. The aims of this research program-still very much in its infancy-- are (1) to elucidate the neuromarker choreography, that is, to determine when each neuromarker is recruited and disengaged, which neuromarkers originate from which brain areas and how neuromarkers interact with each other in transient networks during the course of social behavior. All of the proposed work is geared to the prediction of efficient or deficient outcomes as assessed by detailed single trial analysis of real-time social behavior; (2) to construct a human dynamic clamp that allows for direct manipulation of the interaction between human participants and virtual partners endowed with human appearance and coordinative capacities. This new paradigm opens up the detailed parametric exploration of social behavior; and (3) to integrate the findings in a multiscale neuro- computational model of social behavior, a platform that will enable understanding of basic mechanisms of interpersonal interactions at combined neural, behavioral and social levels. Successful achievement of this program will specify the neurobehavioral routes leading to improved social function. Given the vast number of pathologies with etiological or symptomatic ties to social behavior, such information will afford many translational opportunities for the compensation or remediation of deficits in diseases such as autism, schizophrenia, depression and dementia to name just a few.